Double Data Entry

[EPT]

Hi All,
I have a A97 database that is used primarily for data entry. The
backend is on an NT Server with the front end loaded on data entry
PCs. I would like to do a check on the error rate for entry. Is
there a way to have a second data entry person, re-enter data over
an existing record and have the "system" check to see if the value
the person is entering is the same as that previously entered and if
not an msg would appear? I hope this is clear. I think it would be
similar to opening a previously entered record in edit mode, re-
entering the data but there would be a check between the current
value and the previously entered one. Any thoughts or help would be
appreciated.
TIA
Ed

 

[MichaelRed]

The more traditional approach would be to have the second data entry process populate a second db / table. Assuming there is some key embeded in the data entry process itself, you would then just write a small process to compare the entries. If any are not the same, set a flag in a third table along with the key. If the data does not contain a Key, you need to create one and add it to the source material, to accomplish the record selection for matching.

Your post is somewhat un-clear, as you mention generating a 'error rate' for the dat entry as well as a message (to the 2nd data entry user?). The first (error rate) can only be accomplished after the error is assigned to one of the users entering the data, which implies that the error is resolved between the two data entry processes, while the later implies the second user would be alerted to -and correct- the error.

I have always placed the error resoloution as a third party's responsability, usually a supervisor, making the assignment of the error independent of the actual entry. However I have also noticed a steep decline in the number of requests (and usage) of the redundant entry in recent years. I assume that this is a 'cost cutting' measure based on some sampling of error rates derived from other measures vs the error rate from the redundant entry process.

MichaelRed
m.red@att.net

There is never time to do it right but there is always time to do it over

 

[EPT]

I'll attempt to clarify. Staff are entering surveys, the surveys administered by alcohol/substance abuse treatment program staff were checked for accuracy by two professional staff, so there is little effort in determining the correct entry, simply does it match what is in the data entry box of the survey. In addition, all fields have validation rules. While I can look at error rates for individual staff, I'm primarily interested in the overall error rate. Ultimately, I will be taking a sample of surveys and matching my data analysis file which will be an SPSS file against the original surveys. I've been doing research for a while and haven't had an error rate above .02%. Thanx for all your input.